retro

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs extracting and outputting all "user messages (the human's prompts/questions/instructions)" verbatim into a retrospective, so any secrets pasted by the user in the session would be handled and reproduced by the LLM, creating a high exfiltration risk.